Those of us who work with marketing automation or CRM platforms know the heartbreak of unvalidated data sneaking its way in undetected. Unfortunately, there are a number of ways this can happen. From form submissions and tradeshow lists to manual uploads and acquiring a new company, the quality of your lead data depends on a variety of changing factors.
In an ideal world, when someone fills out a form on your website, they will eventually become a lead in your CRM. This data, however, can undergo a metamorphosis on its journey to Salesforce. Your website picklists may not match up with the values in your MA system, for example, or free form fields make it impossible to control data quality and normalize values.
Fortunately, MA platforms like Eloqua and Marketo provide us with the tools to follow the data from its point of origin to its final destination.
Following the data trail
I like to think of myself and my team as data forensics experts (wouldn’t Data CSI: Finding Out Where Things Took a Wrong Turn make for good television?). It’s really about uncovering the mystery of how bad data values entered your system in the first place, and then mapping back all possible points of egress into a particular contact record or account in your MA platform.
When starting a data cleanup effort, the first thing we’ll do is review all of the activity history for a few sample problem contacts. Most MA platforms provide a history of each contact’s actions, so by following the same path and examining the raw data values you can determine where the data took a wrong turn. Next, we’ll try to duplicate the problem to confirm our findings, much like a mechanic does when you describe that grinding sound coming from under the hood.
Catching the culprit
Once we’ve identified the suspected issue(s), troubleshooting can be as easy as modifying an update rule or updating a field with the corrected data. It can also be as simple as asking the website admin to correct a data value so it enters your MA platform the right way in the first place.
If you have years of historical data, though, the issues can run much deeper. If a free text field was converted into a picklist at some point, you’re probably going to have more data values than you can shake a stick at. By using filters, you can align them with your current picklist values for enhanced segmentation.
It will take time to standardize and map new data values in order to have consistent titles, industries, and so on, but it can be done — and you’ll improve segmentation, scoring, and personalization capabilities as a result.
Some “mistakes” happen for a reason
It’s important that someone pays attention to error emails for when the unexpected occurs. Sometimes, for example, certain Eloqua-to-Salesforce integration events don’t fire correctly. Some errors occur for a reason, though, and it’s important to understand why. For instance, there might be a campaign association that asks for both the lead and contact to fire at the same time. If a contact doesn’t exist yet, the contact campaign association will fail, and that’s perfectly normal.
In fact, I’ve never been in an Eloqua-to-SFDC integration and seen the error log not have some level of errors at any given time. That’s why having the experience to interpret these kinds of errors can be critical to keeping your database healthy.
Don’t let history repeat itself
After you’ve completed your data normalization efforts, it’s critical to put the proper validation in place to prevent the same errors from happening again. Knowing where the bad data originated and understanding how to prevent it from happening again in the future is usually half the battle. It’s also necessary to educate everyone with access to your MarTech platforms on how to enter acceptable data values.
What about those one-off list uploads, where hundreds or thousands of contacts enter your MarTech systems in one fell swoop? One bad list sort in Microsoft Excel can throw off your lead ID values, completely destroying any activity history that could be used for subsequent lead nurturing and scoring programs. As a result, none of the intelligence that is typically recorded in Eloqua makes it into the hands of the business development representatives, and they assume marketing isn’t providing any value.
One way to mitigate uploads directly into your MA system is with a standardized process. By uploading a locked template, along with a concrete set of instructions, users can only choose from approved data values. You can also lock down who can upload a list to a specific set of trusted users. Or, you can have a program running in the background that normalizes the data values behind the scenes each time someone uploads a list.
Seeing the value
After you’ve scrubbed and standardized your data, you can see the value the next time you segment it. The countless hours you’ve lost in the past trying to parse your lists based on wacky data values, or not getting enough contacts into your segments in the first place due to bad data values, will become a distant memory.
What’s more, automated nurture campaigns rely on these more targeted data points. Before, your contacts with bad data points were just sitting in your database gathering dust, never even making it into your campaigns. Now, you can get your message to the right people in the right industry with precision targeting.
Enlist some help
Oftentimes, MA admins have so much on their plate that they just don’t have the time to chase these things down. Many organizations approach us for the express purpose of helping them scrub their data, but it’s also not uncommon for us to uncover data issues when scoping a project for a new client.
I spend a good part of my day making sure data flows correctly, and everyone on my team deals with this in some way, shape, or form almost daily. In fact, we have hundreds of years of experience combined doing this type of analysis. If you’re ready to improve segmentation and give your lead scoring and nurturing programs a boost, we’re here to help.
Bray Coleman, Solutions Architect, designs solutions in marketing automation to help our clients see their campaigns achieve their fullest potential. When he’s not building solutions for our clients you’ll find him chasing his kids, or in the ocean chasing waves.
The post Forensic Data Analysis: Track Down the Bad Data Like a Crime Scene Detective appeared first on DemandGen.
About the Author
More Content by DemandGen
